부산대학교 도서관

Screen-detected (SD) breast cancers are smaller and biologically more indolent than clinically presenting cancers. An often debated question is: if left undiagnosed during their preclinical phase, would they become more aggressive or would they only increase in size? This study considered a registry-based series (1988-1999) of 3329 unifocal, pT1a-pT3 breast cancer cases aged 50-70 years, of which 994 were SD cases and 2335 clinical cases. The rationale was that (1) the average risk of lymph node involvement (N+) is lower for SD cases, (2) nodal status is the product of biological aggressiveness and chronological age of the disease, (3) for any breast cancer, tumour size is an indicator of chronological age, and (4) for SD cases, tumour size is specifically an indicator of the duration of the preclinical phase, that is, an inverse indicator of lead time. The hypothesis was that the relative protection of SD cases from the risk of N+and, thus, their relative biological indolence decrease with increasing tumour size. The odds ratio (OR) estimate of the risk of N+was obtained from a multiple logistic regression model that included terms for detection modality, tumour size category, patient age, histological type, and number of lymph nodes recovered. A term for the detection modality-by-tumour size category interaction was entered, and the OR for the main effect of detection by screening vs clinical diagnosis was calculated. This increased linearly from 0.05 (95%confidence interval: 0.01-0.39) in the 2-7?mm size category to 0.95 (0.64-1.40) in the 18-22?mm category. This trend is compatible with the view that biological aggressiveness of breast cancer increases during the preclinical phase.British Journal of Cancer (2005) 92, 156-161. doi:10.1038/sj.bjc.6602289 www.bjcancer.com Published online 14 December 2004 [ABSTRACT FROM AUTHOR]